Hi all,

I am a medical student attempting to use STATA MP 16.1 to create adjusted KM curves.
I was able to create a graph with two unadjusted KM curves (sts graph, by(THERAPYSEQ)). I then ran a stcox using THERAPYSEQ (the endpoint) and all the confounding variables (13 in total). The unadjusted KM analysis suggests there is a significant difference between the two therapy sequences, but the adjusted cox regression says otherwise, as expected. However, when I use the command "sts graph, by (TNTSEQ) adjust(13 confounding variables)" the graph still shows that the two therapy sequences are significantly different.
Am I using the wrong command? The 13 confounding variables are inside the parentheses, all separated by a space, and their median values are all recoded to 0. I would really like to create a graph with this information for my upcoming poster session, and would like to have clarification that I am presenting accurate data.

Thank you in advance!